Confidence v.s. Critique: A Decomposition of Self-Correction Capability for LLMs

Documenti analoghi

• A Probabilistic Inference Scaling Theory for LLM Self-Correction
  di: Yang, Zhe, et al.
  Pubblicazione: (2025)
• Can Large Language Models Always Solve Easy Problems if They Can Solve Harder Ones?
  di: Yang, Zhe, et al.
  Pubblicazione: (2024)
• Towards Better RL Training Data Utilization via Second-Order Rollout
  di: Yang, Zhe, et al.
  Pubblicazione: (2026)
• Enhancing Reliability across Short and Long-Form QA via Reinforcement Learning
  di: Wang, Yudong, et al.
  Pubblicazione: (2025)
• Exploring Activation Patterns of Parameters in Language Models
  di: Wang, Yudong, et al.
  Pubblicazione: (2024)